PMC text mining subset in BioC: about three million full-text articles and growing

Dataset from a human-in-the-loop approach to identify functionally important protein residues from literature

We present a novel system that leverages curators in the loop to develop a dataset and model for detecting structure features and functional annotations at residue-level from standard publication text. Our approach involves the integration of data from multiple resources, including PDBe, EuropePMC, PubMedCentral, and PubMed, combined with annotation guidelines from UniProt, and LitSuggest and HuggingFace models as tools in the annotation process. A team of seven annotators manually curated ten articles for named entities, which we utilized to train a starting PubmedBert model from HuggingFace. Using a human-in-the-loop annotation system, we iteratively developed the best model with commendable performance metrics of 0.90 for precision, 0.92 for recall, and 0.91 for F1-measure. Our proposed system showcases a successful synergy of machine learning techniques and human expertise in curating a dataset for residue-level functional annotations and protein structure features. The results demonstrate the potential for broader applications in protein research, bridging the gap between advanced machine learning models and the indispensable insights of domain experts.

RegulaTome: a corpus of typed, directed, and signed relations between biomedical entities in the scientific literature

In the field of biomedical text mining, the ability to extract relations from the literature is crucial for advancing both theoretical research and practical applications. There is a notable shortage of corpora designed to enhance the extraction of multiple types of relations, particularly focusing on proteins and protein-containing entities such as complexes and families, as well as chemicals. In this work, we present RegulaTome, a corpus that overcomes the limitations of several existing biomedical relation extraction (RE) corpora, many of which concentrate on single-type relations at the sentence level. RegulaTome stands out by offering 16 961 relations annotated in >2500 documents, making it the most extensive dataset of its kind to date. This corpus is specifically designed to cover a broader spectrum of >40 relation types beyond those traditionally explored, setting a new benchmark in the complexity and depth of biomedical RE tasks. Our corpus both broadens the scope of detected relations and allows for achieving noteworthy accuracy in RE. A transformer-based model trained on this corpus has demonstrated a promising F1-score (66.6%) for a task of this complexity, underscoring the effectiveness of our approach in accurately identifying and categorizing a wide array of biological relations. This achievement highlights RegulaTome's potential to significantly contribute to the development of more sophisticated, efficient, and accurate RE systems to tackle biomedical tasks. Finally, a run of the trained RE system on all PubMed abstracts and PMC Open Access full-text documents resulted in >18 million relations, extracted from the entire biomedical literature.

STRING-ing together protein complexes: corpus and methods for extracting physical protein interactions from the biomedical literature

MOTIVATION

Understanding biological processes relies heavily on curated knowledge of physical interactions between proteins. Yet, a notable gap remains between the information stored in databases of curated knowledge and the plethora of interactions documented in the scientific literature.

RESULTS

To bridge this gap, we introduce ComplexTome, a manually annotated corpus designed to facilitate the development of text-mining methods for the extraction of complex formation relationships among biomedical entities targeting the downstream semantics of the physical interaction subnetwork of the STRING database. This corpus comprises 1287 documents with ∼3500 relationships. We train a novel relation extraction model on this corpus and find that it can highly reliably identify physical protein interactions (F1-score = 82.8%). We additionally enhance the model's capabilities through unsupervised trigger word detection and apply it to extract relations and trigger words for these relations from all open publications in the domain literature. This information has been fully integrated into the latest version of the STRING database.

AVAILABILITY AND IMPLEMENTATION

We provide the corpus, code, and all results produced by the large-scale runs of our systems biomedical on literature via Zenodo https://doi.org/10.5281/zenodo.8139716, Github https://github.com/farmeh/ComplexTome_extraction, and the latest version of STRING database https://string-db.org/.

Improving dictionary-based named entity recognition with deep learning

MOTIVATION

Dictionary-based named entity recognition (NER) allows terms to be detected in a corpus and normalized to biomedical databases and ontologies. However, adaptation to different entity types requires new high-quality dictionaries and associated lists of blocked names for each type. The latter are so far created by identifying cases that cause many false positives through manual inspection of individual names, a process that scales poorly.

RESULTS

In this work, we aim to improve block list s by automatically identifying names to block, based on the context in which they appear. By comparing results of three well-established biomedical NER methods, we generated a dataset of over 12.5 million text spans where the methods agree on the boundaries and type of entity tagged. These were used to generate positive and negative examples of contexts for four entity types (genes, diseases, species, and chemicals), which were used to train a Transformer-based model (BioBERT) to perform entity type classification. Application of the best model (F1-score = 96.7%) allowed us to generate a list of problematic names that should be blocked. Introducing this into our system doubled the size of the previous list of corpus-wide blocked names. In addition, we generated a document-specific list that allows ambiguous names to be blocked in specific documents. These changes boosted text mining precision by ∼5.5% on average, and over 8.5% for chemical and 7.5% for gene names, positively affecting several biological databases utilizing this NER system, like the STRING database, with only a minor drop in recall (0.6%).

AVAILABILITY AND IMPLEMENTATION

All resources are available through Zenodo https://doi.org/10.5281/zenodo.11243139 and GitHub https://doi.org/10.5281/zenodo.10289360.

The biomedical relationship corpus of the BioRED track at the BioCreative VIII challenge and workshop

The automatic recognition of biomedical relationships is an important step in the semantic understanding of the information contained in the unstructured text of the published literature. The BioRED track at BioCreative VIII aimed to foster the development of such methods by providing the participants the BioRED-BC8 corpus, a collection of 1000 PubMed documents manually curated for diseases, gene/proteins, chemicals, cell lines, gene variants, and species, as well as pairwise relationships between them which are disease-gene, chemical-gene, disease-variant, gene-gene, chemical-disease, chemical-chemical, chemical-variant, and variant-variant. Furthermore, relationships are categorized into the following semantic categories: positive correlation, negative correlation, binding, conversion, drug interaction, comparison, cotreatment, and association. Unlike most of the previous publicly available corpora, all relationships are expressed at the document level as opposed to the sentence level, and as such, the entities are normalized to the corresponding concept identifiers of the standardized vocabularies, namely, diseases and chemicals are normalized to MeSH, genes (and proteins) to National Center for Biotechnology Information (NCBI) Gene, species to NCBI Taxonomy, cell lines to Cellosaurus, and gene/protein variants to Single Nucleotide Polymorphism Database. Finally, each annotated relationship is categorized as 'novel' depending on whether it is a novel finding or experimental verification in the publication it is expressed in. This distinction helps differentiate novel findings from other relationships in the same text that provides known facts and/or background knowledge. The BioRED-BC8 corpus uses the previous BioRED corpus of 600 PubMed articles as the training dataset and includes a set of newly published 400 articles to serve as the test data for the challenge. All test articles were manually annotated for the BioCreative VIII challenge by expert biocurators at the National Library of Medicine, using the original annotation guidelines, where each article is doubly annotated in a three-round annotation process until full agreement is reached between all curators. This manuscript details the characteristics of the BioRED-BC8 corpus as a critical resource for biomedical named entity recognition and relation extraction. Using this new resource, we have demonstrated advancements in biomedical text-mining algorithm development. Database URL: https://codalab.lisn.upsaclay.fr/competitions/16381.

RecSOI: recommending research directions using statements of ignorance

The more science advances, the more questions are asked. This compounding growth can make it difficult to keep up with current research directions. Furthermore, this difficulty is exacerbated for junior researchers who enter fields with already large bases of potentially fruitful research avenues. In this paper, we propose a novel task and a recommender system for research directions, RecSOI, that draws from statements of ignorance (SOIs) found in the research literature. By building researchers' profiles based on textual elements, RecSOI generates personalized recommendations of potential research directions tailored to their interests. In addition, RecSOI provides context for the recommended SOIs, so that users can quickly evaluate how relevant the research direction is for them. In this paper, we provide an overview of RecSOI's functioning, implementation, and evaluation, demonstrating its effectiveness in guiding researchers through the vast landscape of potential research directions.

Machine Learning-based Analysis of Publications Funded by the National Institutes of Health's Initial COVID-19 Pandemic Response

Background

The National Institutes of Health (NIH) mobilized more than $4 billion in extramural funding for the COVID-19 pandemic. Assessing the research output from this effort is crucial to understanding how the scientific community leveraged federal funding and responded to this public health crisis.

Methods

NIH-funded COVID-19 grants awarded between January 2020 and December 2021 were identified from NIH Research Portfolio Online Reporting Tools Expenditures and Results using the "COVID-19 Response" filter. PubMed identifications of publications under these grants were collected and the NIH iCite tool was used to determine citation counts and focus (eg, clinical, animal). iCite and the NIH's LitCOVID database were used to identify publications directly related to COVID-19. Publication titles and Medical Subject Heading terms were used as inputs to a machine learning-based model built to identify common topics/themes within the publications.

Results and Conclusions

We evaluated 2401 grants that resulted in 14 654 publications. The majority of these papers were published in peer-reviewed journals, though 483 were published to preprint servers. In total, 2764 (19%) papers were directly related to COVID-19 and generated 252 029 citations. These papers were mostly clinically focused (62%), followed by cell/molecular (32%), and animal focused (6%). Roughly 60% of preprint publications were cell/molecular-focused, compared with 26% of nonpreprint publications. The machine learning-based model identified the top 3 research topics to be clinical trials and outcomes research (8.5% of papers), coronavirus-related heart and lung damage (7.3%), and COVID-19 transmission/epidemiology (7.2%). This study provides key insights regarding how researchers leveraged federal funding to study the COVID-19 pandemic during its initial phase.

Dataset of clinical cases, images, image labels and captions from open access case reports from PubMed Central (1990-2023)

This paper details the acquisition, structure and preprocessing of the MultiCaRe Dataset, a multimodal case report dataset which contains data from 75,382 open access PubMed Central articles spanning the period from 1990 to 2023. The dataset includes 96,428 clinical cases, 135,596 images, and their corresponding labels and captions. Data extraction was performed using different APIs and packages such as Biopython, requests, Beautifulsoup, BioC API for PMC and EuropePMC RESTful API. Image labels were created based on the contents of their corresponding captions, by using Spark NLP for Healthcare and manual annotations. Images were preprocessed with OpenCV in order to remove borders and split figures containing multiple images, data were analyzed and described, and a subset was randomly selected for quality assessment. The dataset's structure allows for seamless integration of different types of data, making it a valuable resource for training or fine-tuning medical language, computer vision or multi-modal models.

Notes on the data quality of bibliographic records from the MEDLINE database

The US National Library of Medicine has created and maintained the PubMed® database, a collection of over 33.8 million records that contain citations and abstracts from the biomedical and life sciences literature. This database is an important resource for researchers and information service providers alike. As part of our work related to the creation of an author graph for coronaviruses, we encountered several data quality issues with records from a curated subset of the PubMed database called MEDLINE. We provide a data quality assessment for records selected from the MEDLINE database and report on several issues ranging from parsing issues (e.g. character encodings and schema definition weaknesses) to low scores for identifiers against several data quality metrics (e.g. completeness, validity and uniqueness). Database URL  https://pubmed.ncbi.nlm.nih.gov.

Europe PMC annotated full-text corpus for gene/proteins, diseases and organisms

Named entity recognition (NER) is a widely used text-mining and natural language processing (NLP) subtask. In recent years, deep learning methods have superseded traditional dictionary- and rule-based NER approaches. A high-quality dataset is essential to fully leverage recent deep learning advancements. While several gold-standard corpora for biomedical entities in abstracts exist, only a few are based on full-text research articles. The Europe PMC literature database routinely annotates Gene/Proteins, Diseases, and Organisms entities. To transition this pipeline from a dictionary-based to a machine learning-based approach, we have developed a human-annotated full-text corpus for these entities, comprising 300 full-text open-access research articles. Over 72,000 mentions of biomedical concepts have been identified within approximately 114,000 sentences. This article describes the corpus and details how to access and reuse this open community resource.

NeuroBridge: a prototype platform for discovery of the long-tail neuroimaging data

Introduction

Open science initiatives have enabled sharing of large amounts of already collected data. However, significant gaps remain regarding how to find appropriate data, including underutilized data that exist in the long tail of science. We demonstrate the NeuroBridge prototype and its ability to search PubMed Central full-text papers for information relevant to neuroimaging data collected from schizophrenia and addiction studies.

Methods

The NeuroBridge architecture contained the following components: (1) Extensible ontology for modeling study metadata: subject population, imaging techniques, and relevant behavioral, cognitive, or clinical data. Details are described in the companion paper in this special issue; (2) A natural-language based document processor that leveraged pre-trained deep-learning models on a small-sample document corpus to establish efficient representations for each article as a collection of machine-recognized ontological terms; (3) Integrated search using ontology-driven similarity to query PubMed Central and NeuroQuery, which provides fMRI activation maps along with PubMed source articles.

Results

The NeuroBridge prototype contains a corpus of 356 papers from 2018 to 2021 describing schizophrenia and addiction neuroimaging studies, of which 186 were annotated with the NeuroBridge ontology. The search portal on the NeuroBridge website https://neurobridges.org/ provides an interactive Query Builder, where the user builds queries by selecting NeuroBridge ontology terms to preserve the ontology tree structure. For each return entry, links to the PubMed abstract as well as to the PMC full-text article, if available, are presented. For each of the returned articles, we provide a list of clinical assessments described in the Section "Methods" of the article. Articles returned from NeuroQuery based on the same search are also presented.

Conclusion

The NeuroBridge prototype combines ontology-based search with natural-language text-mining approaches to demonstrate that papers relevant to a user's research question can be identified. The NeuroBridge prototype takes a first step toward identifying potential neuroimaging data described in full-text papers. Toward the overall goal of discovering "enough data of the right kind," ongoing work includes validating the document processor with a larger corpus, extending the ontology to include detailed imaging data, and extracting information regarding data availability from the returned publications and incorporating XNAT-based neuroimaging databases to enhance data accessibility.

Enhancing thoracic disease detection using chest X-rays from PubMed Central Open Access

Large chest X-rays (CXR) datasets have been collected to train deep learning models to detect thorax pathology on CXR. However, most CXR datasets are from single-center studies and the collected pathologies are often imbalanced. The aim of this study was to automatically construct a public, weakly-labeled CXR database from articles in PubMed Central Open Access (PMC-OA) and to assess model performance on CXR pathology classification by using this database as additional training data. Our framework includes text extraction, CXR pathology verification, subfigure separation, and image modality classification. We have extensively validated the utility of the automatically generated image database on thoracic disease detection tasks, including Hernia, Lung Lesion, Pneumonia, and pneumothorax. We pick these diseases due to their historically poor performance in existing datasets: the NIH-CXR dataset (112,120 CXR) and the MIMIC-CXR dataset (243,324 CXR). We find that classifiers fine-tuned with additional PMC-CXR extracted by the proposed framework consistently and significantly achieved better performance than those without (e.g., Hernia: 0.9335 vs 0.9154; Lung Lesion: 0.7394 vs. 0.7207; Pneumonia: 0.7074 vs. 0.6709; Pneumothorax 0.8185 vs. 0.7517, all in AUC with p< 0.0001) for CXR pathology detection. In contrast to previous approaches that manually submit the medical images to the repository, our framework can automatically collect figures and their accompanied figure legends. Compared to previous studies, the proposed framework improved subfigure segmentation and incorporates our advanced self-developed NLP technique for CXR pathology verification. We hope it complements existing resources and improves our ability to make biomedical image data findable, accessible, interoperable, and reusable.

NLM-Chem-BC7: manually annotated full-text resources for chemical entity annotation and indexing in biomedical articles

The automatic recognition of chemical names and their corresponding database identifiers in biomedical text is an important first step for many downstream text-mining applications. The task is even more challenging when considering the identification of these entities in the article's full text and, furthermore, the identification of candidate substances for that article's metadata [Medical Subject Heading (MeSH) article indexing]. The National Library of Medicine (NLM)-Chem track at BioCreative VII aimed to foster the development of algorithms that can predict with high quality the chemical entities in the biomedical literature and further identify the chemical substances that are candidates for article indexing. As a result of this challenge, the NLM-Chem track produced two comprehensive, manually curated corpora annotated with chemical entities and indexed with chemical substances: the chemical identification corpus and the chemical indexing corpus. The NLM-Chem BioCreative VII (NLM-Chem-BC7) Chemical Identification corpus consists of 204 full-text PubMed Central (PMC) articles, fully annotated for chemical entities by 12 NLM indexers for both span (i.e. named entity recognition) and normalization (i.e. entity linking) using MeSH. This resource was used for the training and testing of the Chemical Identification task to evaluate the accuracy of algorithms in predicting chemicals mentioned in recently published full-text articles. The NLM-Chem-BC7 Chemical Indexing corpus consists of 1333 recently published PMC articles, equipped with chemical substance indexing by manual experts at the NLM. This resource was used for the evaluation of the Chemical Indexing task, which evaluated the accuracy of algorithms in predicting the chemicals that should be indexed, i.e. appear in the listing of MeSH terms for the document. This set was further enriched after the challenge in two ways: (i) 11 NLM indexers manually verified each of the candidate terms appearing in the prediction results of the challenge participants, but not in the MeSH indexing, and the chemical indexing terms appearing in the MeSH indexing list, but not in the prediction results, and (ii) the challenge organizers algorithmically merged the chemical entity annotations in the full text for all predicted chemical entities and used a statistical approach to keep those with the highest degree of confidence. As a result, the NLM-Chem-BC7 Chemical Indexing corpus is a gold-standard corpus for chemical indexing of journal articles and a silver-standard corpus for chemical entity identification in full-text journal articles. Together, these resources are currently the most comprehensive resources for chemical entity recognition, and we demonstrate improvements in the chemical entity recognition algorithms. We detail the characteristics of these novel resources and make them available for the community. Database URL: https://ftp.ncbi.nlm.nih.gov/pub/lu/NLM-Chem-BC7-corpus/.

A reproducible experimental survey on biomedical sentence similarity: A string-based method sets the state of the art

This registered report introduces the largest, and for the first time, reproducible experimental survey on biomedical sentence similarity with the following aims: (1) to elucidate the state of the art of the problem; (2) to solve some reproducibility problems preventing the evaluation of most current methods; (3) to evaluate several unexplored sentence similarity methods; (4) to evaluate for the first time an unexplored benchmark, called Corpus-Transcriptional-Regulation (CTR); (5) to carry out a study on the impact of the pre-processing stages and Named Entity Recognition (NER) tools on the performance of the sentence similarity methods; and finally, (6) to bridge the lack of software and data reproducibility resources for methods and experiments in this line of research. Our reproducible experimental survey is based on a single software platform, which is provided with a detailed reproducibility protocol and dataset as supplementary material to allow the exact replication of all our experiments and results. In addition, we introduce a new aggregated string-based sentence similarity method, called LiBlock, together with eight variants of current ontology-based methods, and a new pre-trained word embedding model trained on the full-text articles in the PMC-BioC corpus. Our experiments show that our novel string-based measure establishes the new state of the art in sentence similarity analysis in the biomedical domain and significantly outperforms all the methods evaluated herein, with the only exception of one ontology-based method. Likewise, our experiments confirm that the pre-processing stages, and the choice of the NER tool for ontology-based methods, have a very significant impact on the performance of the sentence similarity methods. We also detail some drawbacks and limitations of current methods, and highlight the need to refine the current benchmarks. Finally, a notable finding is that our new string-based method significantly outperforms all state-of-the-art Machine Learning (ML) models evaluated herein.

e-TSN: an interactive visual exploration platform for target-disease knowledge mapping from literature

Target discovery and identification processes are driven by the increasing amount of biomedical data. The vast numbers of unstructured texts of biomedical publications provide a rich source of knowledge for drug target discovery research and demand the development of specific algorithms or tools to facilitate finding disease genes and proteins. Text mining is a method that can automatically mine helpful information related to drug target discovery from massive biomedical literature. However, there is a substantial lag between biomedical publications and the subsequent abstraction of information extracted by text mining to databases. The knowledge graph is introduced to integrate heterogeneous biomedical data. Here, we describe e-TSN (Target significance and novelty explorer, http://www.lilab-ecust.cn/etsn/), a knowledge visualization web server integrating the largest database of associations between targets and diseases from the full scientific literature by constructing significance and novelty scoring methods based on bibliometric statistics. The platform aims to visualize target-disease knowledge graphs to assist in prioritizing candidate disease-related proteins. Approved drugs and associated bioactivities for each interested target are also provided to facilitate the visualization of drug-target relationships. In summary, e-TSN is a fast and customizable visualization resource for investigating and analyzing the intricate target-disease networks, which could help researchers understand the mechanisms underlying complex disease phenotypes and improve the drug discovery and development efficiency, especially for the unexpected outbreak of infectious disease pandemics like COVID-19.

Towards a unified search: Improving PubMed retrieval with full text

OBJECTIVE

A significant number of recent articles in PubMed have full text available in PubMed Central®, and the availability of full texts has been consistently growing. However, it is not currently possible for a user to simultaneously query the contents of both databases and receive a single integrated search result. In this study, we investigate how to score full text articles given a multitoken query and how to combine those full text article scores with scores originating from abstracts and achieve an overall improved retrieval performance.

MATERIALS AND METHODS

For scoring full text articles, we propose a method to combine information coming from different sections by converting the traditionally used BM25 scores into log odds ratio scores which can be treated uniformly. We further propose a method that successfully combines scores from two heterogenous retrieval sources - full text articles and abstract only articles - by balancing the contributions of their respective scores through a probabilistic transformation. We use PubMed click data that consists of queries sampled from PubMed user logs along with a subset of retrieved and clicked documents to train the probabilistic functions and to evaluate retrieval effectiveness.

RESULTS AND CONCLUSIONS

Random ranking achieves 0.579 MAP score on our PubMed click data. BM25 ranking on PubMed abstracts improves the MAP by 10.6%. For full text documents, experiments confirm that BM25 section scores are of different value depending on the section type and are not directly comparable. Naïvely using the body text of articles along with abstract text degrades the overall quality of the search. The proposed log odds ratio scores normalize and combine the contributions of occurrences of query tokens in different sections. By including full text where available, we gain another 0.67%, or 7% relative improvement over abstract alone. We find an advantage in the more accurate estimate of the value of BM25 scores depending on the section from which they were produced. Taking the sum of top three section scores performs the best.

Artificial Intelligence-Based Medical Data Mining

Understanding published unstructured textual data using traditional text mining approaches and tools is becoming a challenging issue due to the rapid increase in electronic open-source publications. The application of data mining techniques in the medical sciences is an emerging trend; however, traditional text-mining approaches are insufficient to cope with the current upsurge in the volume of published data. Therefore, artificial intelligence-based text mining tools are being developed and used to process large volumes of data and to explore the hidden features and correlations in the data. This review provides a clear-cut and insightful understanding of how artificial intelligence-based data-mining technology is being used to analyze medical data. We also describe a standard process of data mining based on CRISP-DM (Cross-Industry Standard Process for Data Mining) and the most common tools/libraries available for each step of medical data mining.

Diseases 2.0: a weekly updated database of disease–gene associations from text mining and data integration

The scientific knowledge about which genes are involved in which diseases grows rapidly, which makes it difficult to keep up with new publications and genetics datasets. The DISEASES database aims to provide a comprehensive overview by systematically integrating and assigning confidence scores to evidence for disease–gene associations from curated databases, genome-wide association studies (GWAS) and automatic text mining of the biomedical literature. Here, we present a major update to this resource, which greatly increases the number of associations from all these sources. This is especially true for the text-mined associations, which have increased by at least 9-fold at all confidence cutoffs. We show that this dramatic increase is primarily due to adding full-text articles to the text corpus, secondarily due to improvements to both the disease and gene dictionaries used for named entity recognition, and only to a very small extent due to the growth in number of PubMed abstracts. DISEASES now also makes use of a new GWAS database, Target Illumination by GWAS Analytics, which considerably increased the number of GWAS-derived disease–gene associations. DISEASES itself is also integrated into several other databases and resources, including GeneCards/MalaCards, Pharos/Target Central Resource Database and the Cytoscape stringApp. All data in DISEASES are updated on a weekly basis and is available via a web interface at https://diseases.jensenlab.org, from where it can also be downloaded under open licenses.

Database URL: https://diseases.jensenlab.org

Auto-CORPus: A Natural Language Processing Tool for Standardizing and Reusing Biomedical Literature

To analyse large corpora using machine learning and other Natural Language Processing (NLP) algorithms, the corpora need to be standardized. The BioC format is a community-driven simple data structure for sharing text and annotations, however there is limited access to biomedical literature in BioC format and a lack of bioinformatics tools to convert online publication HTML formats to BioC. We present Auto-CORPus (Automated pipeline for Consistent Outputs from Research Publications), a novel NLP tool for the standardization and conversion of publication HTML and table image files to three convenient machine-interpretable outputs to support biomedical text analytics. Firstly, Auto-CORPus can be configured to convert HTML from various publication sources to BioC. To standardize the description of heterogenous publication sections, the Information Artifact Ontology is used to annotate each section within the BioC output. Secondly, Auto-CORPus transforms publication tables to a JSON format to store, exchange and annotate table data between text analytics systems. The BioC specification does not include a data structure for representing publication table data, so we present a JSON format for sharing table content and metadata. Inline tables within full-text HTML files and linked tables within separate HTML files are processed and converted to machine-interpretable table JSON format. Finally, Auto-CORPus extracts abbreviations declared within publication text and provides an abbreviations JSON output that relates an abbreviation with the full definition. This abbreviation collection supports text mining tasks such as named entity recognition by including abbreviations unique to individual publications that are not contained within standard bio-ontologies and dictionaries. The Auto-CORPus package is freely available with detailed instructions from GitHub at: https://github.com/omicsNLP/Auto-CORPus.

Examining linguistic shifts between preprints and publications

Preprints allow researchers to make their findings available to the scientific community before they have undergone peer review. Studies on preprints within bioRxiv have been largely focused on article metadata and how often these preprints are downloaded, cited, published, and discussed online. A missing element that has yet to be examined is the language contained within the bioRxiv preprint repository. We sought to compare and contrast linguistic features within bioRxiv preprints to published biomedical text as a whole as this is an excellent opportunity to examine how peer review changes these documents. The most prevalent features that changed appear to be associated with typesetting and mentions of supporting information sections or additional files. In addition to text comparison, we created document embeddings derived from a preprint-trained word2vec model. We found that these embeddings are able to parse out different scientific approaches and concepts, link unannotated preprint-peer-reviewed article pairs, and identify journals that publish linguistically similar papers to a given preprint. We also used these embeddings to examine factors associated with the time elapsed between the posting of a first preprint and the appearance of a peer-reviewed publication. We found that preprints with more versions posted and more textual changes took longer to publish. Lastly, we constructed a web application (https://greenelab.github.io/preprint-similarity-search/) that allows users to identify which journals and articles that are most linguistically similar to a bioRxiv or medRxiv preprint as well as observe where the preprint would be positioned within a published article landscape.

HESML: a real-time semantic measures library for the biomedical domain with a reproducible survey

BACKGROUND

Ontology-based semantic similarity measures based on SNOMED-CT, MeSH, and Gene Ontology are being extensively used in many applications in biomedical text mining and genomics respectively, which has encouraged the development of semantic measures libraries based on the aforementioned ontologies. However, current state-of-the-art semantic measures libraries have some performance and scalability drawbacks derived from their ontology representations based on relational databases, or naive in-memory graph representations. Likewise, a recent reproducible survey on word similarity shows that one hybrid IC-based measure which integrates a shortest-path computation sets the state of the art in the family of ontology-based semantic measures. However, the lack of an efficient shortest-path algorithm for their real-time computation prevents both their practical use in any application and the use of any other path-based semantic similarity measure.

RESULTS

To bridge the two aforementioned gaps, this work introduces for the first time an updated version of the HESML Java software library especially designed for the biomedical domain, which implements the most efficient and scalable ontology representation reported in the literature, together with a new method for the approximation of the Dijkstra's algorithm for taxonomies, called Ancestors-based Shortest-Path Length (AncSPL), which allows the real-time computation of any path-based semantic similarity measure.

CONCLUSIONS

We introduce a set of reproducible benchmarks showing that HESML outperforms by several orders of magnitude the current state-of-the-art libraries in the three aforementioned biomedical ontologies, as well as the real-time performance and approximation quality of the new AncSPL shortest-path algorithm. Likewise, we show that AncSPL linearly scales regarding the dimension of the common ancestor subgraph regardless of the ontology size. Path-based measures based on the new AncSPL algorithm are up to six orders of magnitude faster than their exact implementation in large ontologies like SNOMED-CT and GO. Finally, we provide a detailed reproducibility protocol and dataset as supplementary material to allow the exact replication of all our experiments and results.

Software review: The JATSdecoder package-extract metadata, abstract and sectioned text from NISO-JATS coded XML documents; Insights to PubMed central's open access database

JATSdecoder is a general toolbox which facilitates text extraction and analytical tasks on NISO-JATS coded XML documents. Its function JATSdecoder() outputs metadata, the abstract, the sectioned text and reference list as easy selectable elements. One of the biggest repositories for open access full texts covering biology and the medical and health sciences is PubMed Central (PMC), with more than 3.2 million files. This report provides an overview of the PMC document collection processed with JATSdecoder(). The development of extracted tags is displayed for the full corpus over time and in greater detail for some meta tags. Possibilities and limitations for text miners working with scientific literature are outlined. The NISO-JATS-tags are used quite consistently nowadays and allow a reliable extraction of metadata and text elements. International collaborations are more present than ever. There are obvious errors in the date stamps of some documents. Only about half of all articles from 2020 contain at least one author listed with an author identification code. Since many authors share the same name, the identification of person-related content is problematic, especially for authors with Asian names. JATSdecoder() reliably extracts key metadata and text elements from NISO-JATS coded XML files. When combined with the rich, publicly available content within PMCs database, new monitoring and text mining approaches can be carried out easily. Any selection of article subsets should be carefully performed with in- and exclusion criteria on several NISO-JATS tags, as both the subject and keyword tags are used quite inconsistently.

Educating the future generation of researchers: A cross-disciplinary survey of trends in analysis methods

Methods for data analysis in the biomedical, life, and social (BLS) sciences are developing at a rapid pace. At the same time, there is increasing concern that education in quantitative methods is failing to adequately prepare students for contemporary research. These trends have led to calls for educational reform to undergraduate and graduate quantitative research method curricula. We argue that such reform should be based on data-driven insights into within- and cross-disciplinary use of analytic methods. Our survey of peer-reviewed literature analyzed approximately 1.3 million openly available research articles to monitor the cross-disciplinary mentions of analytic methods in the past decade. We applied data-driven text mining analyses to the "Methods" and "Results" sections of a large subset of this corpus to identify trends in analytic method mentions shared across disciplines, as well as those unique to each discipline. We found that the t test, analysis of variance (ANOVA), linear regression, chi-squared test, and other classical statistical methods have been and remain the most mentioned analytic methods in biomedical, life science, and social science research articles. However, mentions of these methods have declined as a percentage of the published literature between 2009 and 2020. On the other hand, multivariate statistical and machine learning approaches, such as artificial neural networks (ANNs), have seen a significant increase in the total share of scientific publications. We also found unique groupings of analytic methods associated with each BLS science discipline, such as the use of structural equation modeling (SEM) in psychology, survival models in oncology, and manifold learning in ecology. We discuss the implications of these findings for education in statistics and research methods, as well as within- and cross-disciplinary collaboration.

PhenoTagger: a hybrid method for phenotype concept recognition using human phenotype ontology

MOTIVATION

Automatic phenotype concept recognition from unstructured text remains a challenging task in biomedical text mining research. Previous works that address the task typically use dictionary-based matching methods, which can achieve high precision but suffer from lower recall. Recently, machine learning-based methods have been proposed to identify biomedical concepts, which can recognize more unseen concept synonyms by automatic feature learning. However, most methods require large corpora of manually annotated data for model training, which is difficult to obtain due to the high cost of human annotation.

RESULTS

In this article, we propose PhenoTagger, a hybrid method that combines both dictionary and machine learning-based methods to recognize Human Phenotype Ontology (HPO) concepts in unstructured biomedical text. We first use all concepts and synonyms in HPO to construct a dictionary, which is then used to automatically build a distantly supervised training dataset for machine learning. Next, a cutting-edge deep learning model is trained to classify each candidate phrase (n-gram from input sentence) into a corresponding concept label. Finally, the dictionary and machine learning-based prediction results are combined for improved performance. Our method is validated with two HPO corpora, and the results show that PhenoTagger compares favorably to previous methods. In addition, to demonstrate the generalizability of our method, we retrained PhenoTagger using the disease ontology MEDIC for disease concept recognition to investigate the effect of training on different ontologies. Experimental results on the NCBI disease corpus show that PhenoTagger without requiring manually annotated training data achieves competitive performance as compared with state-of-the-art supervised methods.

AVAILABILITYAND IMPLEMENTATION

The source code, API information and data for PhenoTagger are freely available at https://github.com/ncbi-nlp/PhenoTagger.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

RENET2: high-performance full-text gene-disease relation extraction with iterative training data expansion

Relation extraction (RE) is a fundamental task for extracting gene–disease associations from biomedical text. Many state-of-the-art tools have limited capacity, as they can extract gene–disease associations only from single sentences or abstract texts. A few studies have explored extracting gene–disease associations from full-text articles, but there exists a large room for improvements. In this work, we propose RENET2, a deep learning-based RE method, which implements Section Filtering and ambiguous relations modeling to extract gene–disease associations from full-text articles. We designed a novel iterative training data expansion strategy to build an annotated full-text dataset to resolve the scarcity of labels on full-text articles. In our experiments, RENET2 achieved an F1-score of 72.13% for extracting gene–disease associations from an annotated full-text dataset, which was 27.22, 30.30, 29.24 and 23.87% higher than BeFree, DTMiner, BioBERT and RENET, respectively. We applied RENET2 to (i) ∼1.89M full-text articles from PubMed Central and found ∼3.72M gene–disease associations; and (ii) the LitCovid articles and ranked the top 15 proteins associated with COVID-19, supported by recent articles. RENET2 is an efficient and accurate method for full-text gene–disease association extraction. The source-code, manually curated abstract/full-text training data, and results of RENET2 are available at GitHub.

Recent advances of automated methods for searching and extracting genomic variant information from biomedical literature

MOTIVATION

To obtain key information for personalized medicine and cancer research, clinicians and researchers in the biomedical field are in great need of searching genomic variant information from the biomedical literature now than ever before. Due to the various written forms of genomic variants, however, it is difficult to locate the right information from the literature when using a general literature search system. To address the difficulty of locating genomic variant information from the literature, researchers have suggested various solutions based on automated literature-mining techniques. There is, however, no study for summarizing and comparing existing tools for genomic variant literature mining in terms of how to search easily for information in the literature on genomic variants.

RESULTS

In this article, we systematically compared currently available genomic variant recognition and normalization tools as well as the literature search engines that adopted these literature-mining techniques. First, we explain the problems that are caused by the use of non-standard formats of genomic variants in the PubMed literature by considering examples from the literature and show the prevalence of the problem. Second, we review literature-mining tools that address the problem by recognizing and normalizing the various forms of genomic variants in the literature and systematically compare them. Third, we present and compare existing literature search engines that are designed for a genomic variant search by using the literature-mining techniques. We expect this work to be helpful for researchers who seek information about genomic variants from the literature, developers who integrate genomic variant information from the literature and beyond.

NLM-Chem, a new resource for chemical entity recognition in PubMed full text literature

Automatically identifying chemical and drug names in scientific publications advances information access for this important class of entities in a variety of biomedical disciplines by enabling improved retrieval and linkage to related concepts. While current methods for tagging chemical entities were developed for the article title and abstract, their performance in the full article text is substantially lower. However, the full text frequently contains more detailed chemical information, such as the properties of chemical compounds, their biological effects and interactions with diseases, genes and other chemicals. We therefore present the NLM-Chem corpus, a full-text resource to support the development and evaluation of automated chemical entity taggers. The NLM-Chem corpus consists of 150 full-text articles, doubly annotated by ten expert NLM indexers, with ~5000 unique chemical name annotations, mapped to ~2000 MeSH identifiers. We also describe a substantially improved chemical entity tagger, with automated annotations for all of PubMed and PMC freely accessible through the PubTator web-based interface and API. The NLM-Chem corpus is freely available.

Protocol for a reproducible experimental survey on biomedical sentence similarity

Measuring semantic similarity between sentences is a significant task in the fields of Natural Language Processing (NLP), Information Retrieval (IR), and biomedical text mining. For this reason, the proposal of sentence similarity methods for the biomedical domain has attracted a lot of attention in recent years. However, most sentence similarity methods and experimental results reported in the biomedical domain cannot be reproduced for multiple reasons as follows: the copying of previous results without confirmation, the lack of source code and data to replicate both methods and experiments, and the lack of a detailed definition of the experimental setup, among others. As a consequence of this reproducibility gap, the state of the problem can be neither elucidated nor new lines of research be soundly set. On the other hand, there are other significant gaps in the literature on biomedical sentence similarity as follows: (1) the evaluation of several unexplored sentence similarity methods which deserve to be studied; (2) the evaluation of an unexplored benchmark on biomedical sentence similarity, called Corpus-Transcriptional-Regulation (CTR); (3) a study on the impact of the pre-processing stage and Named Entity Recognition (NER) tools on the performance of the sentence similarity methods; and finally, (4) the lack of software and data resources for the reproducibility of methods and experiments in this line of research. Identified these open problems, this registered report introduces a detailed experimental setup, together with a categorization of the literature, to develop the largest, updated, and for the first time, reproducible experimental survey on biomedical sentence similarity. Our aforementioned experimental survey will be based on our own software replication and the evaluation of all methods being studied on the same software platform, which will be specially developed for this work, and it will become the first publicly available software library for biomedical sentence similarity. Finally, we will provide a very detailed reproducibility protocol and dataset as supplementary material to allow the exact replication of all our experiments and results.

COVID-19-CT-CXR: A Freely Accessible and Weakly Labeled Chest X-Ray and CT Image Collection on COVID-19 From Biomedical Literature

The latest threat to global health is the COVID-19 outbreak. Although there exist large datasets of chest X-rays (CXR) and computed tomography (CT) scans, few COVID-19 image collections are currently available due to patient privacy. At the same time, there is a rapid growth of COVID-19-relevant articles in the biomedical literature, including those that report findings on radiographs. Here, we present COVID-19-CT-CXR, a public database of COVID-19 CXR and CT images, which are automatically extracted from COVID-19-relevant articles from the PubMed Central Open Access (PMC-OA) Subset. We extracted figures, associated captions, and relevant figure descriptions in the article and separated compound figures into subfigures. Because a large portion of figures in COVID-19 articles are not CXR or CT, we designed a deep-learning model to distinguish them from other figure types and to classify them accordingly. The final database includes 1,327 CT and 263 CXR images (as of May 9, 2020) with their relevant text. To demonstrate the utility of COVID-19-CT-CXR, we conducted four case studies. (1) We show that COVID-19-CT-CXR, when used as additional training data, is able to contribute to improved deep-learning (DL) performance for the classification of COVID-19 and non-COVID-19 CT. (2) We collected CT images of influenza, another common infectious respiratory illness that may present similarly to COVID-19, and fine-tuned a baseline deep neural network to distinguish a diagnosis of COVID-19, influenza, or normal or other types of diseases on CT. (3) We fine-tuned an unsupervised one-class classifier from non-COVID-19 CXR and performed anomaly detection to detect COVID-19 CXR. (4) From text-mined captions and figure descriptions, we compared 15 clinical symptoms and 20 clinical findings of COVID-19 versus those of influenza to demonstrate the disease differences in the scientific publications. Our database is unique, as the figures are retrieved along with relevant text with fine-grained descriptions, and it can be extended easily in the future. We believe that our work is complementary to existing resources and hope that it will contribute to medical image analysis of the COVID-19 pandemic. The dataset, code, and DL models are publicly available at https://github.com/ncbi-nlp/COVID-19-CT-CXR.

TeamTat: a collaborative text annotation tool

Manually annotated data is key to developing text-mining and information-extraction algorithms. However, human annotation requires considerable time, effort and expertise. Given the rapid growth of biomedical literature, it is paramount to build tools that facilitate speed and maintain expert quality. While existing text annotation tools may provide user-friendly interfaces to domain experts, limited support is available for figure display, project management, and multi-user team annotation. In response, we developed TeamTat (https://www.teamtat.org), a web-based annotation tool (local setup available), equipped to manage team annotation projects engagingly and efficiently. TeamTat is a novel tool for managing multi-user, multi-label document annotation, reflecting the entire production life cycle. Project managers can specify annotation schema for entities and relations and select annotator(s) and distribute documents anonymously to prevent bias. Document input format can be plain text, PDF or BioC (uploaded locally or automatically retrieved from PubMed/PMC), and output format is BioC with inline annotations. TeamTat displays figures from the full text for the annotator's convenience. Multiple users can work on the same document independently in their workspaces, and the team manager can track task completion. TeamTat provides corpus quality assessment via inter-annotator agreement statistics, and a user-friendly interface convenient for annotation review and inter-annotator disagreement resolution to improve corpus quality.

SIB Literature Services: RESTful customizable search engines in biomedical literature, enriched with automatically mapped biomedical concepts

Thanks to recent efforts by the text mining community, biocurators have now access to plenty of good tools and Web interfaces for identifying and visualizing biomedical entities in literature. Yet, many of these systems start with a PubMed query, which is limited by strong Boolean constraints. Some semantic search engines exploit entities for Information Retrieval, and/or deliver relevance-based ranked results. Yet, they are not designed for supporting a specific curation workflow, and allow very limited control on the search process. The Swiss Institute of Bioinformatics Literature Services (SIBiLS) provide personalized Information Retrieval in the biological literature. Indeed, SIBiLS allow fully customizable search in semantically enriched contents, based on keywords and/or mapped biomedical entities from a growing set of standardized and legacy vocabularies. The services have been used and favourably evaluated to assist the curation of genes and gene products, by delivering customized literature triage engines to different curation teams. SIBiLS (https://candy.hesge.ch/SIBiLS) are freely accessible via REST APIs and are ready to empower any curation workflow, built on modern technologies scalable with big data: MongoDB and Elasticsearch. They cover MEDLINE and PubMed Central Open Access enriched by nearly 2 billion of mapped biomedical entities, and are daily updated.

FullMeSH: improving large-scale MeSH indexing with full text

MOTIVATION

With the rapidly growing biomedical literature, automatically indexing biomedical articles by Medical Subject Heading (MeSH), namely MeSH indexing, has become increasingly important for facilitating hypothesis generation and knowledge discovery. Over the past years, many large-scale MeSH indexing approaches have been proposed, such as Medical Text Indexer, MeSHLabeler, DeepMeSH and MeSHProbeNet. However, the performance of these methods is hampered by using limited information, i.e. only the title and abstract of biomedical articles.

RESULTS

We propose FullMeSH, a large-scale MeSH indexing method taking advantage of the recent increase in the availability of full text articles. Compared to DeepMeSH and other state-of-the-art methods, FullMeSH has three novelties: (i) Instead of using a full text as a whole, FullMeSH segments it into several sections with their normalized titles in order to distinguish their contributions to the overall performance. (ii) FullMeSH integrates the evidence from different sections in a 'learning to rank' framework by combining the sparse and deep semantic representations. (iii) FullMeSH trains an Attention-based Convolutional Neural Network for each section, which achieves better performance on infrequent MeSH headings. FullMeSH has been developed and empirically trained on the entire set of 1.4 million full-text articles in the PubMed Central Open Access subset. It achieved a Micro F-measure of 66.76% on a test set of 10 000 articles, which was 3.3% and 6.4% higher than DeepMeSH and MeSHLabeler, respectively. Furthermore, FullMeSH demonstrated an average improvement of 4.7% over DeepMeSH for indexing Check Tags, a set of most frequently indexed MeSH headings.

AVAILABILITY AND IMPLEMENTATION

The software is available upon request.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Identification of most influential co-occurring gene suites for gastrointestinal cancer using biomedical literature mining and graph-based influence maximization

Background

Gastrointestinal (GI) cancer including colorectal cancer, gastric cancer, pancreatic cancer, etc., are among the most frequent malignancies diagnosed annually and represent a major public health problem worldwide.

Methods

This paper reports an aided curation pipeline to identify potential influential genes for gastrointestinal cancer. The curation pipeline integrates biomedical literature to identify named entities by Bi-LSTM-CNN-CRF methods. The entities and their associations can be used to construct a graph, and from which we can compute the sets of co-occurring genes that are the most influential based on an influence maximization algorithm.

Results

The sets of co-occurring genes that are the most influential that we discover include RARA - CRBP1, CASP3 - BCL2, BCL2 - CASP3 – CRBP1, RARA - CASP3 – CRBP1, FOXJ1 - RASSF3 - ESR1, FOXJ1 - RASSF1A - ESR1, FOXJ1 - RASSF1A - TNFAIP8 - ESR1. With TCGA and functional and pathway enrichment analysis, we prove the proposed approach works well in the context of gastrointestinal cancer.

Conclusions

Our pipeline that uses text mining to identify objects and relationships to construct a graph and uses graph-based influence maximization to discover the most influential co-occurring genes presents a viable direction to assist knowledge discovery for clinical applications.

PEDL: extracting protein-protein associations using deep language models and distant supervision

Motivation

A significant portion of molecular biology investigates signalling pathways and thus depends on an up-to-date and complete resource of functional protein–protein associations (PPAs) that constitute such pathways. Despite extensive curation efforts, major pathway databases are still notoriously incomplete. Relation extraction can help to gather such pathway information from biomedical publications. Current methods for extracting PPAs typically rely exclusively on rare manually labelled data which severely limits their performance.

Results

We propose PPA Extraction with Deep Language (PEDL), a method for predicting PPAs from text that combines deep language models and distant supervision. Due to the reliance on distant supervision, PEDL has access to an order of magnitude more training data than methods solely relying on manually labelled annotations. We introduce three different datasets for PPA prediction and evaluate PEDL for the two subtasks of predicting PPAs between two proteins, as well as identifying the text spans stating the PPA. We compared PEDL with a recently published state-of-the-art model and found that on average PEDL performs better in both tasks on all three datasets. An expert evaluation demonstrates that PEDL can be used to predict PPAs that are missing from major pathway databases and that it correctly identifies the text spans supporting the PPA.

Availability and implementation

PEDL is freely available at https://github.com/leonweber/pedl. The repository also includes scripts to generate the used datasets and to reproduce the experiments from this article.

Supplementary information

Supplementary data are available at Bioinformatics online.

LitSense: making sense of biomedical literature at sentence level

Literature search is a routine practice for scientific studies as new discoveries build on knowledge from the past. Current tools (e.g. PubMed, PubMed Central), however, generally require significant effort in query formulation and optimization (especially in searching the full-length articles) and do not allow direct retrieval of specific statements, which is key for tasks such as comparing/validating new findings with previous knowledge and performing evidence attribution in biocuration. Thus, we introduce LitSense, which is the first web-based system that specializes in sentence retrieval for biomedical literature. LitSense provides unified access to PubMed and PMC content with over a half-billion sentences in total. Given a query, LitSense returns best-matching sentences using both a traditional term-weighting approach that up-weights sentences that contain more of the rare terms in the user query as well as a novel neural embedding approach that enables the retrieval of semantically relevant results without explicit keyword match. LitSense provides a user-friendly interface that assists its users to quickly browse the returned sentences in context and/or further filter search results by section or publication date. LitSense also employs PubTator to highlight biomedical entities (e.g. gene/proteins) in the sentences for better result visualization. LitSense is freely available at https://www.ncbi.nlm.nih.gov/research/litsense.

PubTator central: automated concept annotation for biomedical full text articles

PubTator Central (https://www.ncbi.nlm.nih.gov/research/pubtator/) is a web service for viewing and retrieving bioconcept annotations in full text biomedical articles. PubTator Central (PTC) provides automated annotations from state-of-the-art text mining systems for genes/proteins, genetic variants, diseases, chemicals, species and cell lines, all available for immediate download. PTC annotates PubMed (29 million abstracts) and the PMC Text Mining subset (3 million full text articles). The new PTC web interface allows users to build full text document collections and visualize concept annotations in each document. Annotations are downloadable in multiple formats (XML, JSON and tab delimited) via the online interface, a RESTful web service and bulk FTP. Improved concept identification systems and a new disambiguation module based on deep learning increase annotation accuracy, and the new server-side architecture is significantly faster. PTC is synchronized with PubMed and PubMed Central, with new articles added daily. The original PubTator service has served annotated abstracts for ∼300 million requests, enabling third-party research in use cases such as biocuration support, gene prioritization, genetic disease analysis, and literature-based knowledge discovery. We demonstrate the full text results in PTC significantly increase biomedical concept coverage and anticipate this expansion will both enhance existing downstream applications and enable new use cases.

Using Manual and Computer-Based Text-Mining to Uncover Research Trends for Apis mellifera

Honey bee research is believed to be influenced dramatically by colony collapse disorder (CCD) and the sequenced genome release in 2006, but this assertion has never been tested. By employing text-mining approaches, research trends were tested by analyzing over 14,000 publications during the period of 1957 to 2017. Quantitatively, the data revealed an exponential growth until 2010 when the number of articles published per year ceased following the trend. Analysis of author-assigned keywords revealed that changes in keywords occurred roughly every decade with the most fundamental change in 1991-1992, instead of 2006. This change might be due to several factors including the research intensification on the Varroa mite. The genome release and CCD had quantitively only minor effects, mainly on honey bee health-related topics post-2006. Further analysis revealed that computational topic modeling can provide potentially hidden information and connections between some topics that might be ignored in author-assigned keywords.

CORD-19: The Covid-19 Open Research Dataset

The Covid-19 Open Research Dataset (CORD-19) is a growing resource of scientific papers on Covid-19 and related historical coronavirus research. CORD-19 is designed to facilitate the development of text mining and information retrieval systems over its rich collection of metadata and structured full text papers. Since its release, CORD-19 has been downloaded over 200K times and has served as the basis of many Covid-19 text mining and discovery systems. In this article, we describe the mechanics of dataset construction, highlighting challenges and key design decisions, provide an overview of how CORD-19 has been used, and describe several shared tasks built around the dataset. We hope this resource will continue to bring together the computing community, biomedical experts, and policy makers in the search for effective treatments and management policies for Covid-19.

CoCoScore: context-aware co-occurrence scoring for text mining applications using distant supervision

MOTIVATION

Information extraction by mining the scientific literature is key to uncovering relations between biomedical entities. Most existing approaches based on natural language processing extract relations from single sentence-level co-mentions, ignoring co-occurrence statistics over the whole corpus. Existing approaches counting entity co-occurrences ignore the textual context of each co-occurrence.

RESULTS

We propose a novel corpus-wide co-occurrence scoring approach to relation extraction that takes the textual context of each co-mention into account. Our method, called CoCoScore, scores the certainty of stating an association for each sentence that co-mentions two entities. CoCoScore is trained using distant supervision based on a gold-standard set of associations between entities of interest. Instead of requiring a manually annotated training corpus, co-mentions are labeled as positives/negatives according to their presence/absence in the gold standard. We show that CoCoScore outperforms previous approaches in identifying human disease-gene and tissue-gene associations as well as in identifying physical and functional protein-protein associations in different species. CoCoScore is a versatile text mining tool to uncover pairwise associations via co-occurrence mining, within and beyond biomedical applications.

AVAILABILITY AND IMPLEMENTATION

CoCoScore is available at: https://github.com/JungeAlexander/cocoscore.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.